stb simulation - local loop of gain-boosting amp

[flyingRabbit]

Hi, below schematic is a fully-differential gain-boosting opamp that I'm working on right now.

I want to check the stability of a local loop that is comprised of gain-boosting amp and a source follower. So, I inserted 'diffstbprobe (red color)' as below, set it as a probe instance, and ran stb simulation (for now, I'm using ideal CMFB, so didn't use pstb).

However, the dc gain of this loop turned out to be 2.8, which is far different from my expectation. I've thought that this dc gain should be close to the gain of boosting amp, which is 114.

Please let me know if you find any mistake or wrong simulation setup.
Thank you in advance.

 

[erikl]

Between source and gate you don't have a phase reversal. So I think you shouldn't reverse the phase at the output of the gain-boosting opamp.

 

[dick_freebird]

stb has let me down enough times that I stick with the old
controlled source (input difference -> single ended mag/phase)
and calculator math. In your case an output difference is
also needed. At least you should have such a testbench
(differing only in whether you use the stb widget, or the
more old-school approach) to determine whether the circuit
is working and stb is bent, or vice versa.

 

[flyingRabbit]

This circuit exploits negative feedback, so reversing phase at the output of the gain-boosting amp is necessary.

 

[flyingRabbit]

Could you explain more details? I used stb because it is the best way of getting loop characteristic, including loading effects of feedback network/reverse transmission, etc.

 

[erikl]

... reversing phase at the output of the gain-boosting amp is necessary.

That's why you probably see such a low loop gain. Doesn't gain-boosting mean you want to boost the gain? That's why you'd have to use positive feedback, isn't it?

 

[flyingRabbit]

That's why you probably see such a low loop gain. Doesn't gain-boosting mean you want to boost the gain? That's why you'd have to use positive feedback, isn't it?

No, erikl. Please refer to https://ieeexplore.ieee.org/xpl/art...queryText=fully differential gain enhancement

 

[Dominik Przyborowski]

The dc OPs of transistors in opamp are correct? Do You use correct CMFB for all opamps?

 

[erikl]

No, erikl. Please refer to https://ieeexplore.ieee.org/xpl/art...queryText=fully differential gain enhancement

I have no access to the IE³ docs, but I believe you anyway: I've misinterpreted the symbols in Razavi's Gain Boosting description
View attachment Razavi__Design_of_Analog_CMOS_ICs_p.311.pdf
Figure 9.26 (a) & (b) because of the missing ring at the amps' outputs, Figure (c) makes the inversion clear, however.

Sorry to muddy the waters!